Risk and Resilience Assessment of Wind Farms Performance in Cold Climate Regions

(1)

Faculty of Science and Technology Department of Technology and Safety

Risk and Resilience Assessment of Wind Farms’

Performance in Cold Climate Regions

Albara Mustafa

A dissertation for the degree of Doctor of Philosophy, September 2022

(2)

i

Risk and Resilience Assessment of Wind Farms Performance in Cold Climate Regions

By Albara Mustafa

A thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy (Ph.D.)

UiT The Arctic University of Norway Faculty of Science and Technology Department of Technology and Safety

(3)

ii

“And it is He (Allah) who sends the winds as good tidings before His mercy”

Quran (25:48)

(4)

iii

(5)

iv

Acknowledgements

This thesis is submitted as a fulfillment of the requirements for the degree of Doctor of Philosophy Ph.D.) at UiT the Arctic University of Norway. The research was carried out in the subject area of Risk and Resilience Assessment of Wind Farms’ Performance in Cold Climate Regions from May 2018 to December 2021.

This work has been carried out in collaboration with my supervisor, Prof. Abbas Barabadi, my co-supervisor, Prof. Yngve Birkelund from UiT the Arctic University of Norway, and my co- supervisor, Prof. Tore Markeset from the University of Stavanger. I wish to thank them for the encouragement and support they provided me with throughout this journey. I would like also to thank Dr. Masoud Naseri for the help he offered me during writing my first paper.

I would like to express my thanks to Bjørn-Morten Batalden, Head of the Department, Javad Barabady, Gunn-Helene Turi, Marith Gabrielsen, Jinmei Lu, for providing me with all necessary facilities and for their valuable guidance and support during my study Many thanks should go to Svein Erik Thyrhaug the manager of Fakken wind farm, and Sonia Lileo from Fortum AS for providing me with wind farms data that was instrumental for completing my work.

I am thankful for the support from all my friends and colleagues at the Faculty of Natural Science and Technology. To name a few, I am grateful to Nikolai Figenschau, Lise Lotte Evenseth, Helene Xue, Hao Chen, Khanh Q. Bui, Sushmit Dhar, Minh Tuan Bui, and all other colleagues at the faculty.

I wish to express my sincere gratitude to my parents, Prof. Mohamad Yazid Mustafa and Eman Fathi, for their support and encouragement, and to my brother Obada, my sisters Isra, Nawar, and Ayah, my wife Karam Taha, and our beloved son Jakob.

Albara Mustafa Tromsø, Norway September 2022

(6)

v

Abstract

Wind energy conversion systems, such as wind farms, are growing in numbers and capacity all over the globe. The onshore wind energy generation sector witnessed an increase of approximately 144 TWh during 2020, with onshore wind farms capacity addition of 108 GW, which is twice as much as the added capacity during 2019 (IEA, 2021). This staggering increase in capacity imposes higher need for improved methodologies and expertise, in measuring and improving the performance of wind farms.

Cold climate regions are known to have an appealing potential for attracting wind farms installation and investments. However, the weather conditions in cold climate regions impose risks and challenges to the operation and maintenance of wind turbines, and to the workers at wind farms. Another challenge prevails in the lack of data and expertise related to wind energy projects in cold climate regions, due to the fact that wind farms installations are relatively new in these regions. Furthermore, cold climate regions are more sensitive to climate changes than other parts of the globe, which increases concerns about the environmental impact of increased investments in wind farms in those regions.

The risks and challenges discussed in this thesis can be classified in different ways, some risks are induced by weather conditions that affect the operation and performance of wind turbines, such as the reliability, availability, and maintainability of wind turbines, and there are the risks that are induced by the wind farms that will affect the societal, the economic, and the environmental status of the surroundings of wind farms.

This thesis introduces applicable methodologies that can be used to measure performance- related aspects of wind farms in cold climate regions, on different levels, and operating under different scenarios. Moreover, in a performance-related context, a methodology for measuring the resilience of wind farms facing disruptive events is introduced, and lastly, the different risks related to the operation of wind farms in cold climate regions are identified and analyzed through a methodology that allows for proper ranking of risks to prioritize the measures that can be used to mitigate those risks.

Keywords

: wind farm; wind turbine; cold climate regions; Arctic region; overall performance index; resilience assessment; risk assessment; operation and maintenance.

(7)

vi

(8)

vii

Abbreviations

WF Wind farm

WT Wind turbine

CCR Cold climate region

MW Megawatt

kW kilowatt

OPI Overall performance index MCDM Multi-criteria decision-making

WSM Weighted sum method

IEC International electrotechnical commission ISO International organization for standardization

NOK Norwegian krone

BN Bayesian network

LCOE Levelized cost of energy CAPEX Capital expenditures OPEX Operational expenditures CA Communication availability CMS Condition monitoring system

SCADA Supervisory Control and Data Acquisition

(9)

viii

Notations

R(t) Reliability

F(t) Probability of failure p Probability of an event

λ Number of events over a specific period, the mean value of the Poisson distribution

t Fixed time interval

k, x Number of events the Poisson distribution finds the probability of

ρ Restoration

R Reliability conditional probability M Maintainability conditional probability S Supportability conditional probability

O Organizational resilience conditional probability

d Throwing distance

D Rotor blade diameter

H Hub height

v Wind speed

Wi Relative weight of performance indicator Si Score of performance indicator

NoisyOrDist Noisy or distribution function NoisyAndDist Noisy and distribution function

Xn Variables in a joint probability distribution μ^A(x) Membership function

X Universal set containing all values of the inputs to the fuzzy logic process

A Fuzzy set

(a, b, c) The three points denoting the triangular fuzzy membership function

(10)

ix

List of appended papers

Paper 1

Mustafa, A. M., A. Barabadi, T. Markeset and M. Naseri (2021), An overall performance index for wind farms: a case study in Norway Arctic region, International Journal of System Assurance Engineering and Management.

Paper 2

Mustafa, A. M. and A. Barabadi (2021), Resilience Assessment of Wind Farms in the Arctic with the Application of Bayesian Networks, Energies, 14(15): 4439.

Paper 3

Albara M. Mustafa, Abbas Barabadi. Criteria-Based Fuzzy Logic Risk Analysis of Wind Farms Operation in Cold Climate Regions. Energies. 2022; 15 (4):1335.

Paper 4

Mustafa, A. M., A. Barabadi and T. Markeset (2019), Risk assessment of wind farm development in ice proven area, Proceedings of the 25 th International Conference on Port and Ocean Engineering under Arctic Conditions (POAC), June 9-13, 2019, Delft, The Netherlands

Paper 5

Mustafa, A., T. Markeset and A. Barabadi (2020), Wind Turbine Failures Review and Gearbox Condition Monitoring, ESREL, Milano, Italy.

Paper 6

Mustafa, A. M., T. Markeset and A. Barabadi (2020), Downtime Cost Estimation: A Wind Farm in the Arctic Case Study, Esrel 2020, Italy.

(11)

x

(12)

xi

Table of Figures

Figure 1. Research process and strategy for this study ... 9

Figure 2. The overall performance model for wind farms ... 10

Figure 3. Unclear visibility at Fakken WF due to snowy weather conditions (Mæhlum, 2013) ... 12

Figure 4. Input and output variables of the resilience of WFs (Mustafa and Barabadi, 2021) 16 Figure 5. Ice types affecting onshore and offshore wind turbines (Mustafa et al., 2019) ... 19

Figure 6. Overall performance index calculation methodology ... 24

Figure 7. Methodology followed to estimate the resilience of WFs using BN. ... 26

Figure 8. An example of a BN with four variables (Mustafa and Barabadi, 2021) ... 27

Figure 8. Graphical depiction of the proposed BN for WF resilience calculation (Mustafa and Barabadi, 2021) ... 28

Figure 9. Methodology of analyzing risks to WFs in CCRs using fuzzy logic ... 29

Figure 10. A triangular membership function ... 31

Figure 11. Membership functions of probabilities (a) and consequences (b) of risks based on experts' judgements ... 32

Figure 12. Risk levels membership functions based on experts' judgements ... 32

Figure 13. Fuzzy risk matrix combining the three variables for risk analysis ... 32

Figure 14. Overview of the Fuzzy logic process ... 33

Figure 15. Relative weights of technical and sustainability performance indicators ... 34

Figure 16. Relative weights of technical and sustainability sub-performance indicators ... 35

Figure 17. relative weight of the availability performance indicators ... 35

Figure 18. Performance indicators scores for Fakken WF ... 36

List of Tables

Table 1: Papers covering research questions ... 3

Table 2: Wind Chill Temperature (WCT) chart (Mustafa and Barabadi, 2022) ... 20

Table 2. A qualitative scale for expressing the OPI ... 25

Table 3. Probabilities ranges for the corresponding probability levels provided by experts ... 31

Table 4. Severity ranges for the corresponding severity levels provided by experts ... 31

Table 5. Risk levels ranges for the corresponding risk levels provided by experts ... 31

Table 6. Risk levels ranges for the corresponding risk levels provided by experts ... 36

Table 7. Summary of calculated resilience for each operating scenario ... 37

Table 8. Enhancement of variables when improving resilience under black swan cold climate conditions ... 38

Table 9. Ranking of risks considering average values of probabilities, consequences, and risk levels ... 38

Table 10. Ranking of risks for the Kozbeyli WF in Turkey using experts’ judgements and Fuzzy logic ... 39

(15)

xiv

(16)

xv

(17)

1

Chapter 1 1 Introduction

1.1 Wind energy in cold climate regions

Wind energy applications are flourishing in cold climate regions (CCRs). CCRs are recognized as regions that experience a minimum hourly temperature at or below -20°C for at least 9 days per year when measurements are taken for a long term such as 10 years or more of measurement (Lehtomäki et al., 2018). Moreover, the long-term average temperature of the location should be below 0°C (Lehtomäki et al., 2018). According to the Global Wind Energy Council, the annual rate of increase of wind energy capacity in CCRs exceeded 20% compared to before 2010 (GWEC, 2011). According to the Wind Power Monthly market update (Lehtomäki, 2016), around 100 Gigawatt (GW) of wind energy were installed in CCRs by the end of 2015.

The same report anticipated that the increase in the installed capacity of wind energy in CCRs would reach 12 GW per year. This significant increase in this sector’s investments should be accompanied by extensive research, to cope with the challenges and risks that might be faced.

Most of the challenges and risks that face wind energy applications installed in CCRs are due to the harsh weather conditions in CCRs. These risks and challenges affect the performance of wind farms (WFs), and their resilience, which is defined as the ability of a technological system to restore its capacity to perform at an acceptable level, when encountering disruptive events (Firesmith, 2019). The majority of current studies concerning the performance of WFs in CCRs focus on the effects of icing on the structural behavior of WTs (Alsabagh et al., 2013), the resulting power losses due to harsh weather conditions (Kilpatrick et al., 2020), the currently used anti/de-icing technologies (Wei et al., 2020) (Dai et al., 2012) (Parent and Ilinca, 2011), and risks caused by ice fall, ice throw, and thrown blade parts (Bredesen and Refsum, 2015) (Rastayesh et al., 2019). What characterizes these studies is that they mostly focus on the technical part of the performance of WFs, at the expense of other performance aspects that concern primarily the risks and challenges caused by the operation of WFs, which might affect the surrounding environment, community, and economy, characterized by the sustainability performance.

Similarly, risks related to WFs in CCRs can be classified into risks caused by harsh weather conditions, such as ice accretion on the blades of WTs, snow accumulation on roads of WFs, and extremely cold temperatures that affect the dexterity of laborers at WFs, and risks caused by the WFs that affect their surrounding environment, community, and economy, such as the impact of WFs on wildlife, the noise and visual annoyance to nearby residents, and their economic situation. The analyses of these risks can provide a holistic view of the different risks that are important for decision-making by designers of WFs, stakeholders, and the community.

Additionally, the performance of WFs under disruptive events, which could be caused by unpredictable extreme weather conditions taking place in CCRs, can be an interesting topic, expressed by the resilience of WFs, which is not researched enough in the literature. There is a lack of a sufficiently comprehensive framework for the resilience of WFs in CCRs, that

(18)

2

discusses resilience from different aspects, which can be an addition to researches such as (Skobiei et al., 2021), where the resilience of offshore WFs is discussed in the light of one factor, which is the redundancy of operating vessels to support the maintenance activities.

Considering several different factors, affecting and shaping the resilience of WFs in CCRs, can provide a more comprehensive framework for measuring resilience, and a demonstration of the interactions between such factors. This can be attained by utilizing the concept of conditional probability as an example, and by using the Bayesian Networks to create different operating scenarios.

The current thesis proposes a hierarchical structure for the performance of WFs, consisting of the technical performance, which can be measured by certain indicators such as the reliability of the wind turbines (WTs), their capacity, and availability. The other part the of performance structure is the sustainability performance, which is concerned with the impacts WFs have on the surrounding environment, community, and economy. Furthermore, the thesis proposes methodologies to calculate the performance of WFs in CCRs, by calculating an overall performance index (OPI). Moreover, the thesis introduces a methodology to measure the resilience of WFs in CCRs under various operating scenarios using the Bayesian networks and analyzes the related risks to WFs in CCRs with the application of Fuzzy logic tools.

1.2 Problem definition

WFs in cold CCRs are subject to several risks that affect their performance. The data and information on the performance and the associated risks to WFs in CCRs might be lacking and insufficient, due to the fact that wind energy applications in such regions are relatively new.

For example, in 2010, the total installed wind energy capacity in Sweden was 2,163 MW, of which only 124 MW was located in cold climate regions. Norway, in the same year, had a total installed wind energy capacity of 436 MW, with only 48 MW installed in cold climate regions (Battisti, 2015). Today, the total wind energy capacity installed in the northern part of Norway, in the counties of Nordland, Troms, and Finnmark, has reached 473 MW, which is nearly 10 times the capacity installed 10 years ago in the same region, the Arctic region. It can be said that the Arctic region and most cold climate regions are among the largest “non-standard”

markets in wind energy today (Lehtomäki et al. 2018).

The newness of the wind energy market in CCRs entails less expertise in the operational conditions that are experienced by WFs, less available data on the operation and maintenance of WTs under severe weather conditions, leading to less comprehensive risk analysis of potential risks, and minimal research on the resilience of WFs in case of unexpected disruptive events taking place.

In order to proceed with a comprehensive assessment of the performance of WFs in the light of potential risks, this study suggests a number of methodologies that can be applied to calculate the overall performance of WFs, identify and analyze the potential risks to and from WFs in CCRs, and calculate the resilience of WFs under various operating scenarios, induced by the weather conditions encountered in the CCRs.

(19)

3

1.3 Purpose and objectives

The purpose of this research is to propose methodologies that can be applied to enhance the performance of WFs in CCRs, and identify the potential risks and challenges that can emerge specifically in CCRs, and affect the performance of WFs, by attaining the following objectives:

1. Developing an index for WFs operating in CCRs that calculates the overall performance of WFs, by combining the technical aspects and the sustainability aspects of WFs.

2. Identifying the most prominent risks related to WFs in CCRs and proposing a methodology to analyze those risks.

3. Assessing the resilience of WFs in CCRs, which can be done by developing a methodology, that calculates the resilience under various scenarios that differ in severity.

1.4 Research questions

In order to fulfill the above-mentioned purpose, three research questions have been formulated in order to help with identifying the key purposes of the research as follows:

Q1. What are the performance aspects that determine the overall performance of WFs operating in CCRs and how can they be measured?

Q2. What are the risks and challenges related to WFs in CCRs during their operation and maintenance and how to analyze them?

Q3. How to measure the resilience of WFs in CCRs under different operating scenarios, including disruptive operating conditions?

These three research questions have been investigated throughout the 6 papers included in this research. The papers attempt to study and discuss the different aspects mentioned in the research questions. Table 1 shows which of the research papers covered and answered which of the research questions.

Table 1: Papers covering research questions

Paper 1 Paper 2 Paper 3 Paper 4 Paper 5 Paper 6

Q1     

Q2     

Q3     

1.5 Scope and limitations

The scope of this research covers the assessment of risks and resilience of the performance of WFs located in CCRs. The research focused on the Arctic region of Norway as a part of the CCRs. The data analyzed in the papers were gathered from WFs located in the Arctic region of Norway. However, the methodologies applied in this research are applicable to WFs located in CCRs or in other non-CCRs.

Regarding the performance of WFs in CCRs, the research is limited to qualitatively calculating an overall performance index for WFs, by measuring two categories of performance indicators:

the technical performance indicators, and the sustainability performance indicators. The

(20)

4

research calculated and compared the overall performance index of a WF located in a CCR to another that is located outside a CCR.

Regarding the resilience of WFs, the research is limited to certain factors contributing to shaping the resilience of wind farms in CCRs. Those factors are the reliability of WTs, the maintainability of the WTs, the supportability of the WF, and the organizational resilience of the WF. Moreover, the research calculates the resilience of WFs as a percentage and was limited to three main operating scenarios, which are the non-cold climate operating conditions scenario, the cold climate operating conditions scenario, and the black swan operating conditions scenario.

Regarding the analysis of risks to WFs, the research is limited to analyzing 6 types of risks, which are i) the increased stoppage rate of WTs due to harsh weather conditions, ii) ice throw from wind turbines, iii) cold stress to workers at wind farms, iv) limited accessibility to wind farms due to snow cover on roads, v) environmental risks caused by the wind farms, and vi) the social opposition risk to installing WFs in CCRs. The research aimed at ranking these 6 risks, depending on their probability of occurrence and severity of consequences, and making a comparison in terms of the ranking of these risks between a WF located in a CCR and a WF located outside this region.

1.6 Data gathering

In order to achieve the goals and objectives of this study, data from two WFs in the Arctic region of Norway were collected throughout this study. Two non-disclosure agreements had to be signed with the two companies owning the two WFs. The gathered WFs data consisted of alarm logs which indicated errors and potential failures the WTs experienced, in addition to the time they were detected and their duration, ice detection events on the blades of the WTs, and the duration of each corresponding stoppage caused by ice accretion, the unavailability of communication events between the WTs and the supervisory system, power production-related conditions such as wind speed, nacelle position, rotor and generator speeds, amount of power produced by each WT, as well as maintenance reports of WTs, which showed the type of failure, the replaced parts, the maintenance activity duration, and the number of personnel carried out the maintenance activities.

In addition, data were collected from experts in the wind energy field, analyzed, and used to make up for the lack of data in certain areas. For example, in Paper 1 experts were asked to assess the relative weight of each performance indicator qualitatively, by giving each indicator a value between (1 and 10), depending on the importance of the indicator to the overall performance of the WF. In Paper 3, experts provided values (from 0 to 10) for the probabilities, consequences, and risk levels in order to plot the membership functions, which were used later in the fuzzy logic process to rank the risks identified in the paper.

(21)

5

1.7 Thesis structure

Chapter 2 presents the background upon which the thesis is built, which encompasses the hierarchical structure of the overall performance of WFs, the resilience of WFs in CCRs, and the identification of WFs-related risks in CCRs. Chapter 3 introduces the research methodologies used to calculate the overall performance index of WFs in CCRs, their resilience under three distinct operating scenarios, and the analysis of six identified risks. The discussion and results of the application of the proposed methodologies, using WFs in the Arctic region of Norway as case studies, are illustrated in Chapter 4. Finally, the conclusions are given in Chapter 5.

(22)

6

1.8 Description of publications Paper 1

Mustafa, A. M., A. Barabadi, T. Markeset and M. Naseri (2021), An overall performance index for wind farms: a case study in Norway Arctic region, International Journal of System Assurance Engineering and Management.

My contribution is developing the methodology used in the paper, communicating with WFs and collecting performance data from one WF, communicating with experts and collecting their answers, analyzing the data, and writing the paper. Masoud Naseri helped me with the calculation made in the paper. Abbas Barabadi and Tore Markeset reviewed the paper and provided comments to improve it.

In Paper 1, we developed a methodology to measure the performance of WFs, by designing a set of performance indicators, that represent the technical and sustainability performance aspects of WFs. Experts in wind energy field provided their assessments regarding the relative weight of each performance indicator. Furthermore, A set of criteria was defined for each performance indicator, and by using the weighted sum method, which is one of the famous methods for multiple-criteria decision making (MCDM), the overall performance index was calculated. This methodology was applied to a WF in the Arctic region of Norway. The resulting overall performance index was 61.3%, which indicated that the WF performance could be described as good. Furthermore, the same methodology was applied to a WF located in a non-cold climate region. Due to the fact that the sustainability performance of this WF was lower than the cold-climate WF, the resulting overall performance index was calculated to be nearly 60%.

Paper 2

Mustafa, A. M. and A. Barabadi (2021), Resilience Assessment of Wind Farms in the Arctic with the Application of Bayesian Networks, Energies, 14(15): 4439.

My contribution is developing the methodology used in the paper, collecting the data of a WF in the Arctic region of Norway, developing and mapping the Bayesian network, running the network, analyzing the data, and writing the paper. Abbas Barabadi reviewed the paper and provided comments to improve it.

In Paper 2, we developed a methodology, using Bayesian networks, to calculate the resilience of WFs as a percentage, while operating under cold climate conditions, and subjected to disruptive events. Three scenarios were defined, and the corresponding resulting resilience was calculated for each scenario. The first scenario implies that the WF operates under non-cold climate conditions, in which the calculated resilience of the WF was the highest. The second scenario is when the WF is operating under cold climate conditions, in which the WF shows a slight degradation in the calculated resilience. The third and final scenario is a black swan scenario, during this scenario the resilience of the WF is significantly reduced due to the severe characteristics of the operating conditions of this scenario.

(23)

7 Paper 3

Albara M. Mustafa, Abbas Barabadi. Criteria-Based Fuzzy Logic Risk Analysis of Wind Farms Operation in Cold Climate Regions. Energies. 2022; 15 (4):1335.

My contribution is developing the methodology of the paper, collecting the data of a WF, collecting and analyzing responses from experts, developing the membership functions and fuzzy logic inference using MATLAB fuzzy logic toolbox, running the model, and writing the paper. Abbas Barabadi reviewed the paper and provided comments to improve it.

In Paper 3 we reviewed the most prominent risks that WFs in cold climate regions are subjected to. In total, 6 risks were identified and analyzed. Experts were communicated to provide their subjective values of probabilities, consequences, and output risk levels (low, medium, high, etc.) for each risk. Afterwards, a set of rules were defined for the different combinations of probabilities and consequences during the fuzzy inference step. A WF in the Arctic region of Norway was selected as a case study, the fuzzy logic toolbox in MATLAB calculated the resulting risk level of all the identified risks for the selected WF, which led eventually to ranking them according to the resulting risks levels. In addition, a WF in a non-cold-climate region was selected to demonstrate the effects of the Arctic operating conditions on the ranking of risks.

Paper 4

Mustafa, A. M., A. Barabadi and T. Markeset (2019), Risk assessment of wind farm development in ice proven area, Proceedings of the 25 th International Conference on Port and Ocean Engineering under Arctic Conditions (POAC), June 9-13, 2019, Delft, The Netherlands

My contribution is reviewing the different types of ice and snow that can accrete on WTs, and the different effects and risks such accretion may represent, during operation and maintenance activities. Moreover, I developed a cross-tabular assessment table that ranks the different types of ice and snow according to the potential risk they may represent to the different parts of onshore and offshore WTs. Abbas Barabadi and Tore Markeset reviewed the paper and provided comments to improve it.

In Paper 4 we outlined the different types of icing that may affect the performance and availability of onshore and offshore WFs located in CCRs. The main types of ice included in the paper are i) Atmospheric icing, ii) Super-structure Icing, and iii) Sea ice. The paper describes the process of formation of each type of ice, and which components in WTs are prone to each ice type. Furthermore, the paper discusses the effects of icing on WTs and WFs in terms of i) mechanical equipment performance, ii) operation and maintenance crew performance, iii) accessibility to WFs and iv) public safety risks. Lastly, the paper proposes a cross-tabular assessment to assess the impacts of icing on the safety of WTs and WFs in CCRs. The conclusion of the paper is that glaze ice and freezing rain and snow induce the highest impact on the structure of WTs. Moreover, operation and maintenance crew performance is highly affected by glaze ice, as it causes slipping, tripping, and falling risks.

(24)

8 Paper 5

Mustafa, A., T. Markeset and A. Barabadi (2020), Wind Turbine Failures Review and Gearbox Condition Monitoring, ESREL, Milano, Italy,

My contribution is reviewing the research done on the measured failure rates of the components of WTs, and the resulting downtime of those failures, in order to determine the most critical component to the availability of WTs. Moreover, I reviewed the commonly used condition monitoring systems (CMS), and Supervisory Control and Data Acquisition (SCADA) systems, to track the health conditions of the different components of WTs. The paper concludes, as other research with a similar aim, that the most critical component to the availability of WTs is the gearbox. Based on that, I reviewed the causes of gearbox failures and the used CM systems to monitor its health condition. Abbas Barabadi and Tore Markeset reviewed the paper and provided comments to improve it.

Paper 5 reviews the critical failures WTs usually experience during their operation by determining the failure rates of components and the resulting downtime from each failure. In addition, the paper provides a brief review of the current CMS and the SCADA systems utilized to monitor the condition and performance of WTs. The paper goes further into Investigating the causes of gearbox failure, which was determined to be the most critical type of failure to WTs, based on the review, and reviews the current CM methods used to monitor the health condition of the gearbox.

Paper 6

Mustafa, A. M., T. Markeset and A. Barabadi (2020), Downtime Cost Estimation: A Wind Farm in the Arctic Case Study, Esrel 2020, Italy,

My contribution is reviewing the contributing costs to the levelized cost of energy (LCOE) from WFs, the risk factors affecting the values of LCOE, selecting a WF in the Arctic region of Norway as a case study, calculating the downtime cost by making use of the LCOE of the WF, caused by the failure of a gearbox in a WT. Abbas Barabadi and Tore Markeset reviewed the paper and provided comments to improve it.

Paper 6 proposes a method to calculate the monetary cost of downtime resulting from a failure in a WT or a WF. The main contributing factors to the LCOE of WFs are the capital expenditures (CAPEX) and the operational expenditures (OPEX). The paper explains the details of each of these contributing factors and how to combine them in an equation to calculate the LCOE. Furthermore, the paper outlines the risk factors that might affect the values of the variables in the LCOE equation. In addition, a WF in the Arctic region of Norway was selected to calculate the monetary losses resulting from the downtime caused by the failure of a gearbox in one of the WTs.

(25)

9

1.9 Research strategy and design

In this research certain aspects related to the performance of WFs, operating in CCRs were covered in six published papers. Figure 1 describes the overall process followed in this research and connects each of the six papers to the process. The process is divided into 3 main activities.

The first activity is to identify the key performance indicators of WFs in CCRs and to calculate the overall performance of WFs by developing an overall performance index, presented mainly in paper 1. The second main activity is to identify the main risks affecting the performance of WFs in CCRs, as well as the main risks WFs induce on their surroundings in CCRs, which were covered and ranked in paper 3. Lastly, the third main activity is to identify the main aspects contributing to the resilience of WFs in CCRs, and to develop a methodology that measures the resilience of WFs when operating under different operational conditions in CCRs, which were covered in Paper 2.

It should be noted that identifying the key risks to WFs in CCRs, in the second main activity of this process, helps with the identification of the most affected WFs performance indicators by the operational conditions in CCRs, described in the first main activity. Moreover, it helps with the identification of the main contributing aspects to the resilience of WFs in CCRs, presented in the third main activity. Therefore, the second main activity in this process can be described as a central activity to the whole research process.

Papers 4, 5, and 6 discussed partially the aspects related to the three main activities. Therefore, the discussions in the following sections of this thesis are mainly focusing on papers 1, 2, and 3 that cover the three main activities extensively.

Figure 1. Research process and strategy for this study

Performance of wind farms (WFs) in cold climate regions (CCRs)

Identify main risks to WFs in CCRs

Identify key aspects of the resilience of WFs in CCRs Identify key performance

aspects to WFs in CCRs

Paper 1 Paper 3 Paper 2

Paper 4

Paper 5 Paper 6

Measure the resilience of WFs in CCRs Rank the risks to and from

WFs in CCRs Measure the overall

performance of WFs in CCRs

(26)

10

Chapter 2 2 Background

2.1 Overall Performance of wind farms in cold climate regions

WFs located in CCRs are subjected to a plethora of challenges. Most of these challenges emerge from the harsh weather conditions such as very low temperatures, ice accretion on the blades of the WTs, and snow accumulation on roads of the WFs, which can hinder the accessibility to the WTs in case they needed maintenance (Lehtomäki et al., 2018). Such challenges affect the technical performance of the WF, which is related to the amount of power produced by the WF (Koo et al., 2018), and can be described and measured by certain indicators, which were developed in this thesis, and described in Figure 2, where the technical performance is constituted by the quality, availability, and capacity performance indicators. In addition, the availability performance indicator can be furtherly sub-categorized into three sub-performance indicators, which are the reliability, maintainability, and supportability performance indicators.

On the other hand, the operation of WFs impacts the surroundings. The impacts can be measured by the sustainability performance of the WF. The social and safety, environmental, and the economic performance indicators represent the three pillars of sustainability performance indicator of a technological system (Diaz-Balteiro et al., 2017).

Figure 2. The overall performance model for wind farms

2.1.1 Technical performance

Technical performance is mainly related to the technical functions of WFs, in terms of the amount of electricity generated (Koo et al., 2018), and how they are affected under cold climate conditions. Technical performance refers to the importance of the quality of the power produced by the WFs, as well as their capacity and availability performances, which can be described in terms of the reliability, maintainability, and supportability of the WFs (IEC, 2015).

2.1.1.1 Quality performance

The quality performance of WFs reflects the design and manufacturing quality of WTs and the WF layout. It also implies maintaining stability between generated and demanded power (Arulampalam et al., 2006). Unstable weather conditions, commonly occurring in CCRs, can cause fluctuations in power production due to significant variations in wind speed. Other

Overall wind farm performance

Technical performance

Quality performance

Availability performance

Reliability performance

Maintainability performance

Supportability performance

Capacity performance

Sustainability performance

Social & Safety impact

Environmental impact

Economic impact

(27)

11

hazards, such as ice accretion on the blades of WTs (Alsabagh et al., 2013) and limited accessibility to WFs due to snow accumulation on the roads (Lehtomäki et al., 2018), will also limit maintenance activities and reduce the quality of power production.

2.1.1.2 Capacity performance

The capacity performance of a WT can be defined as its ability to deliver power according to the design capacity, or according to current demands, in a fixed period with given production resources (Barabady et al. 2010, Shahidul et al. 2013). In light of this definition, the capacity of a WF should reflect the highest sustainable rate of power produced that can be achieved, given the specifications of the WF, the current resources, weather conditions, and maintenance strategies. Capacity can duly affect the efficiency and effectiveness of the operation of a WF (Isaza et al. 2015).

WFs in CCRs are challenged by severe weather conditions, such as ice accretion, snow accumulation, and low temperatures, which can lead to a reduction in wind farm capacity. In addition, the selection of a suitable maintenance strategy plays an important role in attaining the maximum capacity of a WF. Considering these factors, the capacity performance of WFs is expected to be degraded under CCR weather conditions.

2.1.1.3 Availability performance

Availability is defined as “the ability of a functional unit to be in a state to perform a required function under given conditions at a given instant of time or over a given time interval, assuming that the required external resources are provided” (ISO/IEC-2382, 2015). According to the International Electrotechnical Commission (IEC), the availability performance depends upon the combined characteristics of the reliability and maintainability of the item and the maintenance support performance (IEC, 2015), which will be discussed in light of CCR operational conditions.

i) Reliability performance: reliability is defined as “the ability of a component or a system to perform its required functions without failure during a specified time interval, under given conditions”(IEC, 2015). The main aim of system or equipment reliability is to prevent or mitigate the failures that lead to downtimes and reduced power production from WFs.

The rate of failure of WT components may increase under severe weather conditions. Ice and snow may accumulate on the blades of WTs. Snow infiltration inside the nacelle and extreme temperatures may lead to condensation in the electronics and, consequently, can lead to electrical failure (Laakso et al. 2003). For the aforementioned reasons, the blades, control system, and electrical system are responsible for the highest failure rates (Pérez et al., 2013).

Poor component quality of, for example, the variable pitch system, the frequency conversion system, the electrical system, the control system, the gearbox, the generator, and the yaw system can lead to WT breakdown incidents, particularly under harsh weather conditions such as those found in CCRs (Zhang et al., 2013). Moreover, very low temperatures can change the properties of materials and fluids; for example, steel can become more brittle, and lubricants and hydraulic fluids’ viscosity increases (Barabadi and Markeset, 2011).

(28)

12

ii) Maintainability performance: Maintainability is defined as “the ability of an item to be retained in or restored to a state to perform as required, under given conditions of use and maintenance” (IEC, 2015). The maintainability of WTs in CCRs depends to a large extent on the accessibility to the wind farm/turbine to carry out the required maintenance and inspections.

Snow accumulation on the roads of onshore WFs hinders the accessibility to the WTs and calls for snow-removal strategies, or the use of specially equipped vehicles, which will increase the cost of energy (Lehtomäki et al., 2018). Lower temperatures may affect the performance of several materials, such as iron and steel, polymers, and plastics, used in maintenance tools, which experience embitterment at cold temperatures (Markeset et al., 2015). Moreover, maintainability needs to consider human ergonomics, logistics management, design layout, and the level of experience and training of the maintenance personnel (Balindres et al., 2016).

Figure 3 shows an example of the harsh weather conditions maintenance crews experience at Fakken WF, which may hinder proper maintenance activities.

iii) Supportability performance: Supportability is defined as the “ability of an item to be supported to sustain the required availability with a defined operational profile and given logistic and maintenance resources” (IEC, 2015). The supportability of a WF is essentially connected to its maintainability performance, as supportability contributes to fast and frequent maintenance through timely repair/replacement of failed parts in order to maintain the availability of the WF (Kratz 2003). Based on that, numerous factors contribute to the supportability level achieved by WFs. These include logistics considerations of spare parts, personnel, procedures, test equipment, and integrated tools (Smith and Knezevic 1996).

The availability and the location of spare parts have a great impact on the supportability of a product/system (Markeset and Kumar 2005). Spare parts storage at WFs with large-scale WTs is normally limited to small-size spare parts, as it might not be feasible to store large components, such as blades and gearboxes, due to size and capital investment. However, it is the failure of the large-scale components that decreases the availability of WTs significantly and results in the longest downtime, such as the case of the gearbox, which is responsible for almost 56% of the total downtime resulting from failures of WT’s main components (Artigao et al., 2018). Therefore, WF operators tend to order large-scale components from the suppliers once a propagation of failure is observed. In the Arctic region, as an example of a CCR, the

Figure 3. Unclear visibility at Fakken WF due to snowy weather conditions (Mæhlum, 2013)

(29)

13

remote geographical location from suppliers, the cold and harsh climate, and the insufficient and inconvenient infrastructure can affect the effectiveness, and efficiency of the logistics of required supportability services, and the delivery of supplies (Gao and Markeset 2007, Barabadi 2012). In addition, supportability was confirmed, during interviewing experts, to be one of the main challenges to WFs located in remote locations in CCRs.

2.1.2 Sustainability performance

Sustainability science focuses on the management of the relationship between the environment and humans (Afgan et al., 1998), by understanding the interactions between nature and society, meaning that the sustainability goals of a system are achieved through a scientific assessment of the current and the potential future conditions for the Earth System (Omer, 2008).

Sustainability in power production systems implies increasing energy production continuously, using minimum material and energy, as well as non-hazardous materials, cleaning the waste materials resulting from that production in natural ways, decreasing the risks related to human health as far as possible, and using raw materials, including environmental resources, in an efficient way, which in turn results in minimum life-cycle costs (Hallstedt et al. 2010).

The sustainability of WFs located in CCRs should comply with the principles of sustainability, which aim at preserving the ecosystem’s integrity and promoting human health while meeting the demands of the customer and society (Mayyas et al., 2012). Moreover, sustainability implies that WFs should be designed for disassembly, remanufacturing, and recycling, and should be highly recyclable at the end of their life. The conceptual priority in sustainability performance is mainly sustaining society and not explicitly the environment and the economy (Musango and Brent, 2011). Based on this, the sustainability performance indicators of WFs in CCRs are assessed by the following three impacts categories:

• Environmental impacts

• Social impacts

• Economic impacts

2.1.2.1 Environmental impacts of WFs in CCRs

Certain CCRs such as the Arctic are known for their unspoiled nature and wilderness. There are plentiful resources of different fish species, planktonic organisms, and bird habitats, which also make the area vulnerable found in some CCRs. Moreover, it is estimated that the Arctic region might contain 13% of the world’s undiscovered oil and 30% of its undiscovered gas (Gautier et al., 2009). Pollution resulting from energy production from fossil fuels may have serious consequences for the sensitive environment found in CCRs, especially in the Arctic region. WFs, on the other hand, generate electricity carbon-free with no long-term waste, and no cooling water (Pasqualetti, 2011), and are environmentally benign in several ways.

However, their environmental performance needs to be assessed.

Anti-/de-icing chemicals, particularly glycol compounds, such as ethylene, propylene, and alkaline, may be used to de-ice wind-turbine blades, which may create human safety and health

(30)

14

problems, cause environmental harm, represent a threat to surface and groundwater, damage roads and vehicles and may not be cost-effective (Back et al., 1999) and (Dai et al., 2012). In addition, WTs might be one of the reasons for bird mortality. However, research studies stated that, compared to fossil fuels, wind energy killed 20 times fewer birds, and the number of birds killed by WTs may be negligible compared to some other human activities (Sovacool, 2009).

In addition, in the Arctic, as a CCR, WFs might be installed on important winter grazing areas for reindeer, which might lead to changes in reindeers’ density in the region, which might be noticed as well during the construction phase of WFs.

2.1.2.2 Social and safety impact of WFs in CCRs

WFs also have impacts on the surrounding community and its safety. For example, the noise emitted by WTs during their construction and operation, and the visual annoyance that might increase the opposition from the surrounding community to installing WTs in certain areas.

Moreover, ice thrown from operating WTs might be a major concern in CCRs, as pieces of thrown ice might hit the surroundings, including people, cars, animals, and other facilities.

However, this issue might be sometimes exaggerated, as WFs in CCRs are normally located in remote locations, and the severity of icing differs from one WF to another and does not even take place in some WFs, depending on the surrounding geographical and environmental conditions. This was proved during the author’s visits to WFs and by discussing this issue with operators of WFs in northern Norway.

In another context, it can be claimed that governments are violating the rights of indigenous communities by approving wind energy projects in certain areas, causing cultural destruction.

Constructing wind farms on Sámi lands in northern Scandinavia, for example, may be considered unethical and overtly political, simply because it might appear as a systematic dispossession of their lands and a lack of recognition of their rights (Lawrence and Moritz, 2019).

2.1.2.3 Economic impact of WFs in CCRs

Wind energy projects create job opportunities for local communities throughout the wind farm’s lifetime, especially the planning and construction phases since wind energy investments are known to be capital-intensive, with capital costs representing nearly 80% of the total costs of a wind energy project over its lifetime and measured in €/kW (Blanco, 2009). In addition, wind energy promotes the stability of electricity prices in a country, by diversifying the sources of energy. However, most wind energy projects are subsidized by governments due to the high capital and operational expenditures (CAPEX & OPEX) of such projects. Still, technological advances contribute to decreasing these costs, and will eventually lead to a more effective utilization rate of WTs, which is reflected by the percentage of time the WT is operational during the 8760 h (365 × 24) of the year. Thereafter, wind energy projects can yield positive returns on investments. Without even financial support from governments. Moreover, as the prices of fossil fuel-based energy become more expensive, wind energy becomes more competitive.

(31)

15

(32)

16

2.2 Resilience of wind farms in cold climate regions

The resilience of technological systems is defined as the extent to which the system can maintain a certain level of performance when encountered by disruptions (Firesmith, 2019).

WFs in CCRs, as an example of a technological system, are prone to disruptions caused mainly by the harsh weather conditions that affect the resilience of WFs. Such weather conditions create uncertainties about the performance of WFs and how resilient WFs can be in the face of disruptions.

Engineering resilience can be defined mathematically as the sum of reliability and restoration, as per Equation 1 (Youn et al., 2011). Restoration is defined as “the event at which the ‘up’

state is re-established after failure” (IEC, 2015). According to (Rød et al., 2016) restoration depends on several factors, which are (i) the system failure event (i.e. the reliability of the system), (ii) the maintainability of disrupted components, (iii) the supportability of maintenance activities, and (iv) the organizational resilience of the WF.

Resilience (Ψ) = Reliability (R) + Restoration (ρ) (1)

By considering the uncertainties the weather conditions in CCRs may cause to the factors of restoration, a probabilistic approach can be designed to calculate the resilience, which was expressed in Equation 1, as a probabilistic value (between 0 and 1). Therefore, restoration can be expressed as the conditional probability of the previously mentioned four factors, as in Equation 2.

Restoration (ρ) = (1-R) × M × S × O (2)

Where R, M, S, and O are the conditional probabilities of reliability, maintainability, supportability, and organizational resilience respectively. The values of these four variables are conditional, based on the weather conditions WFs experience in CCRs. Therefore, these four factors in addition to restoration and resilience are categorized as probabilistic output variables, with values depending on certain probabilistic input variables, as shown in Figure 4.

WT stoppage Labor dexterity

Snow removal Specialized

vehicles Open roads Spare parts redundancy Communication

availability On-time response

Reliability

Maintainability

Organizational resilience Supportability

Reliability

Restoration Resilience

Figure 4. Input and output variables of the resilience of WFs (Mustafa and Barabadi, 2021)

(33)

17

1. Reliability (R(t)) can be probabilistically expressed as the inverse of the probability of failure (Rausand et al., 2020), as in Equation 3:

R(t) = 1- F(t) (3)

Where F(t) is the probability at which the WTs stop operating due to hazards, caused by the harsh weather conditions in CCRs, or due to component degradation. For simplicity, Poisson distribution is used to represent the probability of the WT stoppage events, as shown in Equation 4 (2007):

p(k;(0,t), λ) = ^{(𝜆𝜆𝜆𝜆)}^𝑘𝑘

𝑘𝑘! 𝑒𝑒^{−𝜆𝜆𝜆𝜆} (4)

Where k is the number of stoppage events of WTs the Poisson distribution finds the probability of, over a fixed period (0, t). λ is the number of WT stoppage events over a specific period, and it represents the mean value of the Poisson distribution.

2. Maintainability. It reflects how easily the system can be maintained, or how quickly the component or the system can be restored to a state, where it can perform at an acceptable capacity (IEC, 2015). The maintainability of WFs in this paper is dependent on two factors, which are the labor dexterity when carrying out the maintenance activities, and the accessibility to the WF, which are both affected by the weather conditions in CCRs.

3. Supportability. The supportability activities are tightly connected to the maintainability of WFs. Supportability to WFs in CCRs are mainly concerned with the on-site availability and the provision of WT spare parts, and maintenance tools that will help the service team to restore the performance of the WF and its availability, during and after disruptive events. Therefore, as shown in Figure 4, supportability is dependent on the redundancy of spare parts, and the accessibility to public roads to deliver the needed parts and tools from suppliers to the WF site.

4. Organizational resilience. The resilience of a WF as an organization implies the capacity of the operational team to prepare for disruptive events, respond, and adapt to them, whether these disruptive events take place gradually or as sudden (BS-65000, 2014).

Therefore, the probabilistic approach to measuring the organizational resilience of WFs in CCRs depends on:

• Communication availability (CA), which encompasses the communication between the operational team and the WF through monitoring systems. Incidents that lead to loss of data gathered from WTs, through SCADA systems and condition monitoring systems, render the communication with the WF unavailable. A Poisson distribution is used to estimate the probability of loss of connection events (x), taking place over a specific period (0, t), with considering an average number of loss-of-connection incidents (λ). Hence, the probability of connection availability can be represented as per Equation 5 (Zio, 2007):

CA = 1-p(x;(0,t), λ) =1- ^{(𝜆𝜆𝜆𝜆)}^𝑥𝑥

𝑥𝑥! 𝑒𝑒^{−𝜆𝜆𝜆𝜆} (5)

• On-time response to events. It represents how successful the response of the WF operator to the disruptive events the WF encounters, which can be also assessed by the percentage of times the operator takes action to respond to the disruptive event,

(34)

18

which can be described as an on-time response. For example, if 85% or more of the disruptive events are responded to by the WF operator within the first hour of their occurrence, then the WF operator can be described as resilient, and the on-time response variable can be considered 100% successful (Hosseini and Barker, 2016).

(35)

19

2.3 Wind farms-related risks in cold climate regions

Paper 3 differentiates between risks caused by the harsh weather conditions in CCRs that affect the technical performance of WFs, and risks caused by the WFs, which affect the surroundings of the WFs such as the environment and the nearby community. Based on that, six types of risks are analyzed, which are as follows:

2.3.1 Risks caused by weather conditions that affect the performance of WFs

The following risks affect the technical performance of WFs in terms of their reliability, maintainability, and supportability performances described earlier in Fig. 1, which in turn affect the availability performance of WFs and their power production. These risks are as follows:

1. Increased WT stoppages due to harsh weather conditions (WT stoppage). This risk en- compasses the stoppages WTs experience that is caused by the harsh weather conditions, which affect the WTs and lead to increasing their stoppage rate in different ways.

The physical properties of materials are affected by low temperatures in CCRs. For example, the gearbox lubricating oil viscosity differs with variation in temperatures, when the temperatures are very low, the viscosity of lubricating oil increases, and flows more slowly, creating more friction and thus negatively impacting the efficiency of the gearbox by overheating it and higher fatigue charges (Laakso et al., 2005).

In addition, the ice accretion on the blades of the WTs leads to increased load on the structure of the WT, and imbalanced and unsafe operation, leading to shutting down the WT to avoid major losses such as losing the WT. As a consequence of that, the power production from the WT will be lost until the accreted ice melts down and the operation of the WT is restored (Andersen et al., 2011). Figure 5 illustrates the different types of ice that accrete on different parts of both onshore and offshore WTs. These ice types are furtherly explained and discussed in Paper 4.

The two most common types of ice that accrete on the blades of onshore WTs are rime ice and glaze ice. Rime ice forms when supercooled water droplets freeze immediately upon impacting the surface of the WT blade, while glaze ice forms when the liquid water freezes shortly after impacting the surface of the blade (Bravo Jimenez, 2018).

Glaze ice accretion forms near the freezing point (0 ^oC) and has strong adhesion to the surface, it is transparent and has a higher density than rime ice. On the other hand, rime

Icing types

Atmospheric icing

In cloud icing

Glaze

Hard rime

Soft rime

Precipitation

Freezing rain

Wet snow

Frost Super-

structure icing Sea ice

Land-fast

Floating

Figure 5. Ice types affecting onshore and offshore wind turbines (Mustafa et al., 2019)

(36)

20

ice has lower adhesion to the surface and has a white or opaque color, and can be easily removed compared to glaze ice (Xue and Khawaja, 2016).

2. Cold stress to workers (Cold stress). Cold temperatures cause cold stress to crew work- ers and limit their dexterity (Wærø et al., 2018). Serious cold-related illnesses and injuries, caused by trench foot, frostbite, and hypothermia, may occur in case of extremely cold temperatures, in addition to permanent tissue damage, and death that may result as a consequence of major cold-related injuries (Mustafa and Barabadi, 2022).

High wind speeds and cold temperatures are the two main factors contributing to cold stress for workers (Osczevski and Bluestein, 2005). Wind Chill Temperature (WCT) is a measure that determines the likelihood that workers are subjected to the risk of frostbite, which can be calculated using Eqn. 6, where V is the wind speed (km/h) 10 m above the surface and T is the air temperature (°C) (Osczevski and Bluestein, 2005):

𝑊𝑊𝑊𝑊𝑊𝑊[°𝑊𝑊] = 13.12 + 0.621𝑊𝑊 −11.37𝑉𝑉^0.16+ 0.3965𝑊𝑊𝑉𝑉^0.16 (6)

Table 2 has been generated using Eqn. 6. The table is used to determine whether the workers at WFs in CCRs are subjected to the risk of frostbite or not, where the shaded region indicates an increased risk of frostbite (Osczevski and Bluestein, 2005).

Table 2: Wind Chill Temperature (WCT) chart (Mustafa and Barabadi, 2022) Air Temperature (°C)

10 5 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50

Wind Speed (km/h)

10 9 3 -3 -9 -15 -21 -27 -33 -39 -45 -51 -57 -63 15 8 2 -4 -11 -17 -23 -29 -35 -41 -48 -54 -60 -66 20 7 1 -5 -12 -18 -24 -31 -37 -43 -49 -56 -62 -68 25 7 1 -6 -12 -19 -25 -32 -38 -45 -51 -57 -64 -70 30 7 0 -7 -13 -19 -26 -33 -39 -46 -52 -59 -65 -72 35 6 0 -7 -14 -20 -27 -33 -40 -47 -53 -60 -66 -73 40 6 -1 -7 -14 -21 -27 -34 -41 -48 -54 -61 -68 -74 45 6 -1 -8 -15 -21 -28 -35 -42 -48 -55 -62 -69 -75 50 6 -1 -8 -15 -22 -29 -35 -42 -49 -56 -63 -70 -76 55 5 -2 -9 -15 -22 -29 -36 -43 -50 -57 -63 -70 -77 60 5 -2 -9 -16 -23 -30 -37 -43 -50 -57 -64 -71 -78 70 5 -2 -9 -16 -23 -30 -37 -44 -51 -59 -66 -73 -80 80 4 -3 -10 -17 -24 -31 -38 -45 -52 -60 -67 -74 -81

3. Limited accessibility to wind farms due to snow cover on the roads. This risk is primarily related to the maintenance of WTs, as accumulated snow on the roads of a WF might

Risk and Resilience Assessment of Wind Farms Performance in Cold Climate Regions

Risk and Resilience Assessment of Wind Farms’

Performance in Cold Climate Regions

Risk and Resilience Assessment of Wind Farms Performance in Cold Climate Regions

Acknowledgements

Abstract

Keywords

Abbreviations

Notations

List of appended papers

Table of Contents

Table of Figures

List of Tables

Chapter 1

1 Introduction

Chapter 2

2 Background